A slightly different approach to deck-building.

[Captain Scibra]

I actually use this theory for making my decks sometimes 

Yeah, I'm quite aware that many people use it and have used it for a long time, this is just putting it into an understandable manner such that people that don't use it can learn to use it for future building of decks. 

...sounds like the get it when you need it approach...

Yes, it is.

Also if there are more questions, I'll be glad to answer, for I may have left a few holes in the theory.

[coinich]

Heh, it suggested I take .9 less Pillars and .2 more Eclipses...

[Tea is good]

weird I thought of the same thing, except I based it around making a three turn win. Mine has too many variables, but it was based off a normally constructed deck I was using. It felt very, very similar to play

Your deck strat

Code: [Select]

4vj 4vj 4vj 6u2 7ae 7ae 7ae 7dp 7dp 7dp 7dq 7dq 7dq 7dq 7dq 7dq 7jv 7jv 7jv 7qb 7tb 7tb 7tb 809 809 809 809 809 809 80k
reg deck strat

Code: [Select]

4vj 4vj 4vj 4vj 6u2 74a 77g 77g 7ae 7dp 7dp 7dp 7dp 7dp 7dq 7dq 7dq 7dq 7dq 7dq 7jv 7qb 7tb 809 809 809 809 809 809 80k

[QuantumT]

The idea is somewhat limited, as it is assumes that the cards in your deck are evenly distributed, which isn't always going to be the case. However it will give you a general idea.

The reason the smallest possible deck is desirable is that the possibility of poor draws is lower. Let's take your deck for instance. A 32 card, 4 otyugh version has a 1 % better chance of having that second turn otyugh. A single percent might not sound like much, but there's no cost for this improvement. It also applies to every card in your deck.

It's also worth noting that as the game continues, the percentages grow. The chance to have at least 2 otys by turn 8 is 2% higher for the 32 card deck (also of note is the fact that the chance of having exactly 2 otyughs is actually 3% higher).

This is taking it to an extreme but I think it helps illustrate what is going on. Let's say you're comparing a 60 card deck with 2 eternities to a 30 card deck with 1 eternity. The chance of the 30 card deck not drawing its eternity in 30 cards is 0 (obviously). The chance for the 60 card deck is roughly 25%.

Long story short, unless you're a stall or have card drawing, the smaller the better.

[Captain Scibra]

Heh, it suggested I take .9 less Pillars and .2 more Eclipses...

Well the idea is that the number given is really a minimum value, meaning extra copies are accepted and actually probably suggested, due to the "give or take" of random drawing, but it's not entirely necessary.

weird I thought of the same thing, except I based it around making a three turn win. Mine has too many variables, but it was based off a normally constructed deck I was using. It felt very, very similar to play

Your deck strat

Code: [Select]

4vj 4vj 4vj 6u2 7ae 7ae 7ae 7dp 7dp 7dp 7dq 7dq 7dq 7dq 7dq 7dq 7jv 7jv 7jv 7qb 7tb 7tb 7tb 809 809 809 809 809 809 80k
reg deck strat

Code: [Select]

4vj 4vj 4vj 4vj 6u2 74a 77g 77g 7ae 7dp 7dp 7dp 7dp 7dp 7dq 7dq 7dq 7dq 7dq 7dq 7jv 7qb 7tb 809 809 809 809 809 809 80k

In the case of Novae and Cremation decks, as I failed to mention their unique form of quanta gain, you can group cards from different elements that serve the same general function, so substitutes can be made.  In other words, as in the regular strategy, an Improved Blessing can be replaced with Unstoppable.  As for the damage cards, Giant Frogs and Vampire Daggers can be replaced by Elite Graboids, also in the regular strategy.  I'm not exactly sure how you got that many damage cards in the strategy using my theory.  I loosely implied at the end of my example that after balancing key cards, it's cleanup.  Fixing QI, in your case adding more Novae and Cremations.  Then for the "give or take" of random drawing, add more key cards.  Lastly damage, unless of course the key function of the deck is concentrating on fast damage.  I'm not sure if I answered your question entirely, since you never really gave one directly.

The idea is somewhat limited, as it is assumes that the cards in your deck are evenly distributed, which isn't always going to be the case. However it will give you a general idea.

The reason the smallest possible deck is desirable is that the possibility of poor draws is lower. Let's take your deck for instance. A 32 card, 4 otyugh version has a 1 % better chance of having that second turn otyugh. A single percent might not sound like much, but there's no cost for this improvement. It also applies to every card in your deck.

It's also worth noting that as the game continues, the percentages grow. The chance to have at least 2 otys by turn 8 is 2% higher for the 32 card deck (also of note is the fact that the chance of having exactly 2 otyughs is actually 3% higher).

This is taking it to an extreme but I think it helps illustrate what is going on. Let's say you're comparing a 60 card deck with 2 eternities to a 30 card deck with 1 eternity. The chance of the 30 card deck not drawing its eternity in 30 cards is 0 (obviously). The chance for the 60 card deck is roughly 25%.

Long story short, unless you're a stall or have card drawing, the smaller the better.

Well first off, the deck used as the main example is a low-damage stall, so I found 40 to be a comfortable decksize.  And as mentioned above, adding extra copies is acceptable to make up for that "give or take" with random drawing.  Even if the "give or take" is not considered, the theory reduces the maximum of each end since the theoretical build is based off producing the ideal outcome. 

[QuantumT]

Well first off, the deck used as the main example is a low-damage stall, so I found 40 to be a comfortable decksize.  And as mentioned above, adding extra copies is acceptable to make up for that "give or take" with random drawing.  Even if the "give or take" is not considered, the theory reduces the maximum of each end since the theoretical build is based off producing the ideal outcome. 

I use the word stall a bit differently than most. When I say stall I'm referring to a deck that intends to win via deckout. I would classify this as a mid-range control deck.

However, the my main point is the idea that "adding extra copies is acceptable to make up for that "give or take" with random drawing" is fallacious. Adding more copies won't make up for the give or take with random drawing, it will only make it worse.

And I don't think it's a good idea to assume the ideal outcome, as games are won and lost based on what happens in non ideal draws. This method will get you in the ballpark though.

[Captain Scibra]

I use the word stall a bit differently than most. When I say stall I'm referring to a deck that intends to win via deckout. I would classify this as a mid-range control deck.

However, the my main point is the idea that "adding extra copies is acceptable to make up for that "give or take" with random drawing" is fallacious. Adding more copies won't make up for the give or take with random drawing, it will only make it worse.

And I don't think it's a good idea to assume the ideal outcome, as games are won and lost based on what happens in non ideal draws. This method will get you in the ballpark though.

Ok I can understand your interpretation of stall versus it just being a really slow but balanced deck.

On the note about how adding extra copies makes the "give or take" worse, I completely disagree.  You are calculating the minimum number of a specific card required for it to theoretically appear so often and the ratio is linked directly with the decksize, which should never change during the whole process.  When you add extra copies you would use the leftover space that was not filled by the balanced number of each key card.  In other words, you saying that adding extra copies of key cards seems to imply that adding other cards as support for the deck is even worse at putting the decks main strategy into effect. 

[coinich]

If we're looking at absolute probabilities, look at my sig.  Hopefully it will clear up any issues in that regard.

[QuantumT]

On the note about how adding extra copies makes the "give or take" worse, I completely disagree.  You are calculating the minimum number of a specific card required for it to theoretically appear so often and the ratio is linked directly with the decksize, which should never change during the whole process.  When you add extra copies you would use the leftover space that was not filled by the balanced number of each key card.  In other words, you saying that adding extra copies of key cards seems to imply that adding other cards as support for the deck is even worse at putting the decks main strategy into effect.

Let me illustrate what I'm talking about with an example that's much simpler than elements.

In this example, we'll use standard playing cards. Your goal in this game is to get 1 card of each color when you draw a pair. Obviously the ideal "deck" contains an even split of red and black cards. Let's look at decks with different numbers. Here the first number refers to how many of each card are in the deck, and the next is the ratio of games that you win.

1-1
2-2/3
3-3/5
4-4/7
5-5/9
6-6/11
etc.

The more of each card you have, the higher the chance to fail. It's exactly the same in elements.

Those numbers I gave you aren't just something I made up either. Those differences are actual, calculated probabilities, and the math doesn't lie. It is fact that the bigger you make the deck, the more prone it becomes to bad, uneven draws.

If you're going to add completely different cards that another thing entirely, but it doesn't change the fact that the deck should be as small as possible.

[Tea is good]

I made the second deck on my first post first. It was a DEMA, except I wanted more speed. Then I saw that a certain combo of cards (1 cremation, 2 BoL, 1 nova, 2 Lava Destroyers, 1 vamp dag and 1 blessing) could give me a three turn win. So I decided to make a deck, centered around getting these cards, in their portions in my first hand (8 cards on a missed flip). That was the first deck on the list.

[Captain Scibra]

On the note about how adding extra copies makes the "give or take" worse, I completely disagree.  You are calculating the minimum number of a specific card required for it to theoretically appear so often and the ratio is linked directly with the decksize, which should never change during the whole process.  When you add extra copies you would use the leftover space that was not filled by the balanced number of each key card.  In other words, you saying that adding extra copies of key cards seems to imply that adding other cards as support for the deck is even worse at putting the decks main strategy into effect.

Let me illustrate what I'm talking about with an example that's much simpler than elements.

In this example, we'll use standard playing cards. Your goal in this game is to get 1 card of each color when you draw a pair. Obviously the ideal "deck" contains an even split of red and black cards. Let's look at decks with different numbers. Here the first number refers to how many of each card are in the deck, and the next is the ratio of games that you win.

1-1
2-2/3
3-3/5
4-4/7
5-5/9
6-6/11
etc.

The more of each card you have, the higher the chance to fail. It's exactly the same in elements.

Those numbers I gave you aren't just something I made up either. Those differences are actual, calculated probabilities, and the math doesn't lie. It is fact that the bigger you make the deck, the more prone it becomes to bad, uneven draws.

If you're going to add completely different cards that another thing entirely, but it doesn't change the fact that the deck should be as small as possible.

Ok I think I see what you're getting at then.  Just reread the whole back-and-forth and I had missed your point in the first post when I had first read it, for I may not have been concentrating that well.  Yes a smaller decksize does give less bad drawing, but having a 30 card deck isn't always ideal.  In many cases, best example being a deckout stall, having a larger decksize compliments the strategy in that it may be pretty close to a 30-cards-drawn game.  Like my deck example, a win usually happens after more than 20 cards have been drawn when against opponents who have a decent amount of creature control, so damage can be hard to keep out, since the use of 2 elements is rather limiting, and that damage isn't the main strategy in the first place.  Hence forth it will generally take longer to kill the opponent.  Having less than 10 cards often near the end of the game really doesn't sit well with me, which is why I felt 40 was a comfortable decksize with the field control strategy.  So having a smaller deck to decrease the bad draws and make a card more likely to appear by the time its needed isn't always the solution.

[Captain Scibra]

I made the second deck on my first post first. It was a DEMA, except I wanted more speed. Then I saw that a certain combo of cards (1 cremation, 2 BoL, 1 nova, 2 Lava Destroyers, 1 vamp dag and 1 blessing) could give me a three turn win. So I decided to make a deck, centered around getting these cards, in their portions in my first hand (8 cards on a missed flip). That was the first deck on the list.

So the fact that you didn't substitute for the multiple Blessings and had less Novae and Cremations, how did it actually affect the deck?

It felt very, very similar to play

Here you say it felt similar but you never mentioned whether it was a positive or negative change in how the deck performed.  I doubt it was positive, but since you tested them, I shall leave that up to you to say.